LE is not a usable metric for the quality of a health care system, for reasons understandable and verifiable by any with a minimum of statistical/casuality analysis background. Your own sources say so. Repeatedly. The only metrics that are capable of valid proxy use are those assessing treatment outcomes for specific conditions. Even these cannot give a comprehensive picture unless aggregated properly, which no one to date seems to have done.

In other words your argument is that health care is impossible to measure.

I disagree.

And if you disagree with that statement you have to use life expectancy. Period.

Nobody will ever “properly” aggregate the data. Most sources I can access say we’re best on cancer survival rates, but mediocre on the death rates. This means one of those two numbers is total BS.

I’m assuming the cause is that a) comparative studies focus on a few cancers; whereas Wikipedia lists almost 200:
http://en.wikipedia.org/wiki/List_of_cancer_types

BTW, your source wasn’t very good. It was a study of less than a half-dozen forms of cancer, using data from 1999. And it used 5-year survival rate.

If you’d tell me what general area of the country you live in, I will be happy to provide local/regional references to clinical personnel who can assist you with that profound reading comprehension problem you exhibit. Or were you misrepresenting what I’ve said on purpose?

Do I have to call Justin on you you silly person?

Which of these statements is untrue:

1) You believe health costs will sink us.

2) You believe that eventually we will be forced into a government-dominated system to get them under control.

3) You oppose this inevitability because it will hamper some abstract thing you call “freedom,” and increase another abstract thing you call “tyranny.”

In other words you’d prefer costs to go up, which necessarily leads to people dieing, to paying taxes because taxes are anti-freedom.

BTW, we are already more anti-freedom by your definition than the Brits. They pay less tax money for health care per capita than we do, but they still have access to private medicine. Maybe you should study that OECD page some more before you talk about how bad government control is.

Nick, all you’ve done is make claims for LE as a generalized proxy metric for health care system quality that are completely unsupported and are contradicted by the very sources you cite. LE is not a usable metric for the quality of a health care system, for reasons understandable and verifiable by any with a minimum of statistical/casuality analysis background. Your own sources say so. Repeatedly. The only metrics that are capable of valid proxy use are those assessing treatment outcomes for specific conditions. Even these cannot give a comprehensive picture unless aggregated properly, which no one to date seems to have done.

Sorry, but I’m not able to cure willful ignorance. The light bulb’s gotta wanna change — picking metrics because they suit what you want to believe does not magically make them valid metrics. Yes, that’s willful base ignorance on your part.

Dude, you are claiming that sources you have not cited, say that in general we do better on cancer than anyone else.

Since you’re apparently unequipped to locate any valid sources of condition-specific cross-national data (perhaps you can’t access google) here’s some cites of the most recent published studies of cross-national cancer survival rate comparisons. (From non-U.S. sources, I note, just to proactively head off your inevitable claim of biased nationalism.) If you lack the resources to access the full text of the 2008 Lancet Oncology study, or the intellectual capacity to comprehend the data therein, here’s a handy chart of some of the areas covered. The Telegraph article on the 2007 study already has non-specific charts on overall cancer survival rates, and yes, the LO authors did adjust for the things you think are gameable.

The reasons for this disparity do indeed involve structural/cultural differences in HC systems and utilization, particulary in greater population screening, earlier detection, and aggressive treatment.

And apparently Tully dislikes the idea of universal care because he’d prefer to see costs for everybody spiral out of control (as he admits they will probably do without single-payer), to paying for that fat chick’s diabetes meds.

If you’d tell me what general area of the country you live in, I will be happy to provide local/regional references to clinical personnel who can assist you with that profound reading comprehension problem you exhibit. Or were you misrepresenting what I’ve said on purpose?

Here, I’ll try to keep this brief, though I’ve already said it all above, but seem to have stated it in a fashion above your comprehension level and limited attention span. ALL health care systems in ALL developed nations that are not tyrannies or dictatorships exhibit and suffer from the excess cost growth problem, regardless of their form of health care system, whether it’s single-payer like the UK or blended like the US. In addition, observation indicates that the greater the amount of state control of the health care system, the more nanny-state anti-freedom tyrannical behaviors based on health care justifications said state tends to exhibit.

Free states have excess cost growth problems because people in general demand more additional health care for each additional dollar in income. Period. Demand for health care is price-inelastic. BTW, the US is not even high among OECD countries in excess cost growth. We’re right about average or a hair under. Since you have trouble finding such data, I’ll provide this handy pointer to the OECD web site of frequently requested health care data.

Or you can remain in willful base ignorance, arguing based on your own opinions and desires, rather than on the facts.

Problem being that I’m not the one cherry-picking, Nick. That’s your straw man.

Dude, you are claiming that sources you have not cited, say that in general we do better on cancer than anyone else.

In terms of general health there are many important measures that count besides cancer. You picked one that supports your argument. That’s a cherry. Which you picked, rather than picking the death rate from preventable disease.

And according to an actual data source:
http://www.nationmaster.com/graph/hea_dea_fro_can-health-death-from-cancer
Our cancer death rate is mediocre. So you not only picked a specific subset of diseases (Cancer) that looked good for us, you picked a specific measure (survival rates for particular cancers) that looked good. Apparently the pro-US bias in those survival rates is bigger than I thought.

You gonna have to work a little bit harder than that to win this point.

And as I said I never claimed Life Expectancy was perfect. I just claimed that it was a decent measure of a health system’s overall effectiveness because it actually measures overall health. OTOH all you got is a vague claim that we do better in one specific set of diseases that most people never get (cancer). Which is apparently not true, because the French, Scandinavians, Brits, and Aussies have lower cancer death rates.

As for your “rationing” argument, it’s indicative of a base ignorance of the field. The mere existence of privatized elements in the UK does not negate the fact that most Brits get nothing but government care, and that said care is 90% or more of their national HC expenditures, or that the taxation to fund same effectively prevents all but the upper classes from accessing any substantial care outside the NHS.

“Base ignorance?” You try to redefine words and accuse me of “base ignorance?” And then you claim a literal interpretation of your “real definition” shows ignorance? Come on now.

Heck by your argument we already ration to a huge chunk of Americans — the elderly largely can’t afford non-Medicare insurance, homeless vets can’t afford anything but the VA, Medicare recipiants (by definition) can;t afford anything but Medicare, and at any given time 10-20% of Americans are without healthcare because they can’t afford anything. And if middle-class Brits are rationed because they can’t afford private health care (a point many of them would dispute, especially given that less of their tax money, per capita, goes to health care), it’s pretty clear a huge chunks of Americans are rationed too.

@WHQ
You didn’t miss much.

An NC Doctor told a fat chick she was doomed because she was too fat, and got in trouble because he poked her when he did it.

The anecdote was used to illustrate the point that under universal health care we’d all pay for that fat chicks bad habits.

And apparently Tully dislikes the idea of universal care because he’d prefer to see costs for everybody spiral out of control (as he admits they will probably do without single-payer), to paying for that fat chick’s diabetes meds.

Not really, because they still have some choice on obtaining their insurance and on seeking outside services if their insurance doesn’t cover what they want. If you’re on Medicaid, you by definition entirely lack the financial ability to seek any other option. In the UK, if you’re not at least solidly middle class, the amount of taxes paid for NHS has pretty much swallowed the funding you would need to seek outside services. If you’re middle class or better, you can of course join the other Brits who travel to other nations to get medical care rather than waiting through the queue-rationing of NHS. Just because you theoretically pay for services in such a system as the NHS doesn’t mean you will ever get them.

Many Americans also travel to get treatments at cheaper cost. And many insurance companies will pay for that — even with the travel costs to, say, India, it’s still cheaper for them to go there for a bypass than to get one here. American health care is indeed shockingly expensive by any standard.

I thought you were an advocate of a single-payer system a few years back

I did frequently state that the way trends were running, that’s where we would probably eventually end up. And I did frequently say that as far as “bending the cost curve,” that was probably the only way to actually control excess cost growth. (See above re: excess cost growth and rationing.) That’s not remotely the same as advocacy. I am opposed to nationalized government-controlled health care for what I consider sound reasons, this one among them.

It’s essentially the same as claiming that Medicaid recipients really don’t have any rationing being imposed on them because they can always go to the Mayo Clinic and pay for it themselves.

Couldn’t the same be said about most Americans with private insurance? Isn’t really a matter of what is or is not covered by your insurance unless you’re wealthy enough to cover your health-care expenses out of pocket? Or are you discussing the degree of rationing rather than the existence of it?

Side bar: I thought you were an advocate of a single-payer system a few years back, Tully, during the Centerfield heyday. Is that right, or did you simply provide lots of information about single-payer in health-care discussions without actually advocating it?

Problem being that I’m not the one cherry-picking, Nick. That’s your straw man. And apparently you didn’t bother to read the paper you’re citing other than to cherry-pick, or you might have noticed they say much the same things I have above about the usefullness of LE and LE-derived measures as a proxy measure for HC quality. Not that the Robert Wood Johnson Foundation is an impartial-peer-review-publication organization (it isn’t, and has a definite agenda) or that the authors relied extensively on early-to-mid 90′s stats themselves (they did, which seems somewhat odd for a 2009 paper claiming in the title to be a “Timely Analysis of Immediate Health Policy Issues,” and purportedly evaluating quality in a sector characterized by high rates of technological change) but here, since you insist, from the paper you tried to cherry-pick:

An important issue in health care quality measurement, as in other types of research that attempt to ascertain causality, is that it is very difficult to adjust for factors outside the health care system which contribute to particular health outcomes, such as socioeconomic status, lifestyle, and disease incidence or prevalence.

Got that? Clear enough? And specifically referring to the use of mortality/LE-derived measures as HC-system quality measures:

Although very interesting as indicators of health status, these [both derived and generalized mortality/LE measures] all fall short as measures of health care quality because they tend to be significantly influenced by factors other than health care.

IOW, I don’t think that paper says what you think it says. It does, however, note some of the areas where we excel and some where we suck. Or rather, where we excelled and sucked fifteen years ago or so.

As for your “rationing” argument, it’s indicative of a base ignorance of the field. The mere existence of privatized elements in the UK does not negate the fact that most Brits get nothing but government care, and that said care is 90% or more of their national HC expenditures, or that the taxation to fund same effectively prevents all but the upper classes from accessing any substantial care outside the NHS. NHS rations, and they have little choice unless they have much higher incomes than average. It’s essentially the same as claiming that Medicaid recipients really don’t have any rationing being imposed on them because they can always go to the Mayo Clinic and pay for it themselves.

It’s not compelling to know that you doubt that other countries’ lifestyles are healthier than ours. And what do you count as lifestyle? Do crime rates, occupations and wars fall under lifestyle?

I’m talking about industrialized nations. None of those have experienced a truly major war since Vietnam, and we were the only industrialized nation that actually took significant casualties. None have been occupied since the Soviet Empire fell.

By lifestyle I mean the biggies — weight/diet/exercise, alcohol, cigarettes, and drugs. We are clearly fatter than anybody else, which implies our diets are worse and/or we get less exercise. But we do very well in per capita cigarette and alcohol consumption. Greece’s cigarette consumption is 3,000 per person per year:
http://en.wikipedia.org/wiki/List_of_countries_by_cigarette_consumption_per_capita
And they still beat us in life expectancy from birth. I can’t find data for life expectancy at 5, tho.

Other lifestyle choices don’t produce statistics (weed consumption per capita data is hard to find), or only affect a tiny slice of the population. For example our homicide rate is 5.8 per 100,000:
http://en.wikipedia.org/wiki/List_of_countries_by_intentional_homicide_rate
That’s .000058% of Americans. That’s a tragedy, but it’s not gonna account for two full years of life expectancy. Especially life expectancy at 50.

When there are multiple factors involved in determining life expectancy, which specific, narrow factor can I not point to as the reason for the differences in life expectancies among industrialized nations by suggesting that the aggregate of the other factors can’t be that different among those industrialized nations? Why can’t I point to any given factor “X” and say that all the other factors must even out, so factor “X” must be the cause?

You can, but the problem is there just aren’t that many other factors between industrialized countries. We fight more wars, and have more crime, but even if we were the only country in the world to do those things it still wouldn’t knock years off the average person’s life. We are excessively fat, but the Canadians are close behind, and I have yet to find an industrialized country that doesn’t smoke or drink more than we do. What else would you use?

Regardless of that, the issue here is life expectancy versus alternatives. And alternatives all have their own flaws. The cancer survival rate’s Tully’s using would be of limited value even if they weren’t a) cherry-picked, and b) artificially inflated by early discovery. You could try something like death rate, but that would be biased against the Europe because European countries are much older than us.

As far as the British system being rationed, the government system is, indeed, rationed by the government. But there’s no law against a purely private practice. That means that, by your “real meaning” definition the Brits do not have rationing. If the private market there is too small for your taste that’s an indication that almost nobody who actually has to live with NHS rationing is so unhappy with it that they go to private hospitals.

As for treatment outcomes,
You’re cherry-picking. I can cherry-pick too:
http://www.rwjf.org/qualityequality/product.jsp?id=47508
We suck at stomach-cancer and hip fractures. Apparently our advantage with most other cancers is not statistically significant. In terms of overall care in 10 studies these guys reviewed comparing quality in the US to Canada Canada won five, and three did not show statistically significant differences. That’ right the second-crappiest health system in the industrialized world went 5-2 against us.

The problem with your argument about cost growth is that it treats health care as just another sector of the economy. In economic theory that works great. In reality I have never met a person who gets more utility from $999,999,999,999 then they do from their spouse. In other tech sectors choices are easier — a $999,999,999,999 computer just won’t sell.

When there are multiple factors involved in determining life expectancy, which specific, narrow factor can I not point to as the reason for the differences in life expectancies among industrialized nations by suggesting that the aggregate of the other factors can’t be that different among those industrialized nations? Why can’t I point to any given factor “X” and say that all the other factors must even out, so factor “X” must be the cause?

For my part, I don’t have any interest in proving any particular approach is better. Instead, I’m interested in trying a variety of approaches, hopefully finding a convergence of data. Nick seems pretty interested in plumping life expectancy and demeaning treatment outcomes. Data doesn’t lie iof you understand its limits and pay close attention to it. But people can use it to mislead, and that’s always a problem.

The reason I push Life expectancy at five is it is simple, easily measured, very difficult to game, and measures the health outcome people care about the most: life.

Treatment rates are great when you’re discussing treatments, but none of the health bills we’re talking about was written with specific treatments in mind. They were written with the entire system in mind, and I have yet to see a statistic that sums up the entire system better then life expectancy. It’s possible the number is bad for some reason (ie: maybe American fatasses are harder to keep alive then Irish drunkards), but IMO it’s highly unlikely that every Industrialized country has an overall healthier lifestyle than we do.

I sincerely doubt our lifestyle is so much unhealthier then theirs that they can get measurably better results (higher life expectancy) overall while spending roughly half what we do.

You’re missing the larger point, which is that the generalized LE stat is completely useless as a proxy for HC system quality in a developed nation. It simply does not measure what it is being purported to measure. What very little correlation there is between aggregate LE and HC system Q is completely drowned out by other more highly-correlated factors. Its only real utility is as an intra-national indicator, assuming that all other relevant factors have been adjusted for — itself a monumental task and near-insurmountable obstacle. As an international comparitive it’s essentially useless due to multiplication of much stronger factors. So why use it that way at all? (Only one reason: to rationalize/justify pre-determined political positions.)

Nick continues to try to dismiss what utility LE does have in specific-outcome analysis through his “stat-gaming” argument, but as I said, a review of specific-condition treatment outcomes pretty well destroys that case. There is nothing terribly ambiguous about a positive biopsy for malignancy or metastatic-stage diagnosis of same, and the stats for specific conditions and specific treatment outcomes are easily rendered robust and usable for cross-national outcomes comparisons for specific conditions. We do indeed outperform the world in many specific areas, mostly due to three factors. Earlier detection, leading-edge treatment, and quicker commencement of treatment after detection.

Those better outcomes are the direct result of the higher price we pay for health care. As is the shrinking affordability of them. Pain, gain.

Which, as you say, can only be solved via some form of rationing. There’s a real conflict between the desire for as much treatment as one wants and the growing need for costs to be manageable across the socioeconomic spectrum, right?

Not exactly. While economists speak of “market rationing,” it’s not really rationing at all, but the aggregated outcome of free choices made by individuals achieving a dynamic equilibrium. Real rationing is the controlled distribution of scarce resources, goods, or services, REGARDLESS OF market equilibrium. REAL rationing is the only way to limit the “excess” cost growth of HC. But that cost growth is “excessive” only in the sense that it grows faster than GDP. Health care is hardly the only sector to exhibit such growth over the ages. Consider, for example, the growth in the %GDP cost of the transportation sector for the years 1900-1950 … did we see a need for government rationing of cars because that sector grew faster than GDP? How about computing services and hardware over the last half-century? We gonna ration that? Heh.

The excess cost growth problem for tech sectors (and health care is assuredly a tech sector) is common to ALL developed nations, including the vaunted single-payer nations that ALREADY have de facto rationing, such as the UK. Trying to avoid that inherent market-demand pressure by assigning the rationing decisions to government just changes the pressure from market pressure to purely political pressure. In anything resembling a democratic system the result is the same — excess cost growth — as market demand is never allowed to reach an equilibrium point. I do not see that as a feature of governmental control, and outside of a purely tyrannical statist system it demonstrably does not “solve” the “problem.” We pay increasing amounts of our incremental next-dollar income for health care because as both individuals and a society we want it, pure and simple. The demand is in the price-inelastic range, and until it reaches the elastic range excess cost growth will continue.

There are no magic wands here. No silver bullets.

Indeed. The only reason for pursuing that nice average seems to be to make specious generalist claims, one way or another. At least that’s the only way it seems to actually be used, isn’t it? There does not appear to be any other actual utility involved, does there? So why use questionable proxies at all, if actual comparisons on specifics are available, and have actual utility?

All good points except the specious part. The general proxy can indeed be used to make specious claims if you’re intent upon debating only to support a preconceived point of view. Anyone who wants to make specious claims can do so by selective use of any data, right?

For my part, I don’t have any interest in proving any particular approach is better. Instead, I’m interested in trying a variety of approaches, hopefully finding a convergence of data. Nick seems pretty interested in plumping life expectancy and demeaning treatment outcomes. Data doesn’t lie iof you understand its limits and pay close attention to it. But people can use it to mislead, and that’s always a problem.

I think you’re doing a good job of pointing out the various catch-22s in healthcare. And I’ve been listening to you speak to various aspects of it for several years now. So it doesn’t trouble me that in this particular post you’re speaking to the issue of individual choice. Things you’ve told me on other occasions make me comfortable that you understand that individual choice, desirable though it may be, can in fact contribute negatively to the basic problem, which you define as a problem of limited supply and unlimited demand. Which, as you say, can only be solved via some form of rationing. There’s a real conflict between the desire for as much treatment as one wants and the growing need for costs to be manageable across the socioeconomic spectrum, right?

We both know that as costs continue to rise comparably to the rates of the past several decades, the price will become more and more untenable. Since we live in a republic that functions via representative democracy, we both know that if the prices (and dislocations) become more untenable for more people, voters will opt for an approach that pools risk. Leaving aside for a moment any question of moral rectitude or economic efficiency, I think we would, if pushed, agree that the tea leaves suggest that healthcare in America is going to move closer towards some sort of single-payer/single adminstrator approach. Whether we like it or not. Absent some astonishing economic turnaround, I can’t foresee any big turn towards freer markets and multiplying private provision. Do you?

That’s why I am hoping that the system can evolve to allow for a variety of choices within a unified system. People who want more protection and more treatment should pay higher premiums. People who want less should pay less. But if they decide later that they want more, there should be some premium for upgrading.

And people should be able to have options that include the ability to keep their costs down by agreeing to pay more of the cost of actual treatment. So for example, suppose you hurt your wrist. And the protocol is not to get a CAT scan for the best image of your injured wrist until other approaches (x-rays, rest, immobilization, cortisone) have failed. If you want to skip ahead to GO and get that CAT scan right away, you pay more of the cost than if you waited. Maybe you get one “skip ahead to Go” option per year. I dunno, point is, such things can be managed rationally, even if the government is involved.

Some want to take those sorts of very personal decisions away from insurance companies, and place them in the hands of patients and their doctors. But that involves patients and their doctors taking over the system, doesn’t it? And placing the good of “patients” over the good (profits) of the insurance companies? Hmmm. What do we call health-care systems like that?

The patient may well (and usually does) have a somewhat different viewpoint about said cost-benefit, because the benefit is THEIR life being saved if they hit that early-detection lottery against the overall cost-benefit “book odds.”

The other side of this, though, is the role of doctors in over-testing. Not all of it is patient-driven, and even some that is gets plenty of help from the doctor. My personal opinion is that there has to be a push for better stardards for best-practices among physicians. How many and sever of symptoms for a brain tumor does someone have to exhibit before you, as the physician, decide to send them for a CT scan (layered x-ray, btw)? Some people are going to ask for tests, and I say, if they do, inform them of the risks of testing versus not, and let them have them if they still want them. But physicians should be assessing the risks/benefits on a purely rational basis and shouldn’t be sending people, say, with headaches, and nothing else indicitive of brain problems, off for CT scans unless the patient insists despite the doctor’s advice. People should have the choice. I’m fuzzy on whether or not someone else, including an insurer, should pay for it in every case.

Life expectancy is a truly shitty proxy measure of health care quality, systemic or otherwise. HC quality has only a marginal effect on LE stats in developed countries, as Frank notes. The way we eat, how many miles we drive, the high propensity of young inner-city males to kill each other … all of these have a greater impact on LE stats than HC system quality.

As you pointed out simply using results from any specific disease is worse because those vary significantly across populations. Life expectancy has several major advantages:

1) It is not easily gamed. Long-term survival rates can be artificially increased simply by doing lots more tests. If you catch Breast Cancer two years before the the other guys do you’ve got two free years in the long-term survival rates game. If you’re an asshole it’s even easier — it’s not like a woman with a lump in her breast knows whether it’s cancerous, so a Doctor who finds lots of lumps and declare them all cancer looks like a miracle worker because most of his patients live decades after diagnosis. OTOH you know when your mom died.

2) It is not subject to over-analysis. Different populations have different medical problems, and deciding which diseases should be counted (and how much they should count) is not a trivial task. For example if you use lots of the diseases fat people get our numbers are gonna suck because we’re fattest. But if you use liver diseases the Irish, and Czechs drink a lot more beer than we do:
http://en.wikipedia.org/wiki/List_of_countries_by_alcohol_consumption
For the Czechs it’s almost double (BTW: Luxembourg is more than double on this list, but it’s a tiny country with low Liquor taxes surrounded by big countries with much higher liquor taxes, so a lot of their alcohol is actually consumed by foreigners.

It’s true you’ve got to factor in things like infant mortality, which are measured slightly differently in different countries. That’s why I used life expectancy at five. His point about inner-city kids killing each other is less relevant, but I checked US and French life expectancy at 50 anyway. Ours was almost 81, theirs was more than 83.

Tully’s absolutely right when he says we’re fatasses. But we’re also teetotalers. France is #6 on that alcohol consumption list, we’re #43. Greece is worse than us in terms of smoking and drinking, and it’s a much poorer country to boot, but their life expectancy at birth is better than ours (I had to use birth, Wolfram Alpha doesn’t have data for them at 5 or 50).

IMO fattasses that spend almost 50% more than drunks do on health care should expect to live at least as long as the drunks. And we spend 16% of GDP on health care, while the French only spend 11.2%. In absolute terms the disparity is worse — they spent $3,600 per person we spend $7,100.